00:00I do want to talk about where we stand in this AI cycle.
00:03There's been so much talk, certainly in financial markets, about digital intelligence in the AI build-out.
00:09But over the last few weeks, we've seen Jensen Wong very publicly, from his travels in Korea and Taiwan and
00:16beyond, talk a lot more about physical AI.
00:20And you made a big bet on that about two years ago when you agreed to buy ANSYS for about
00:25$35 billion.
00:26A high value, a high premium. Why?
00:29Because we see the same thing.
00:31I mean, AI started with the digital AI.
00:35Think of it when you move to physical AI, you're moving the agentic brain to operate a physical system, be
00:42it a car, a robot, a drone.
00:45You need that intelligence to merge.
00:47The difference in the physical AI, it operates in the wild, in an uncontrolled environment.
00:53You need to model the environment.
00:55You need to simulate the environment.
00:57In order to do it in a cost-effective way, you need to create a digital twin of both the
01:03product and the environment in order to reduce the cost of development and safety, security, and other aspects of operating
01:11the product in the physical world.
01:13It took you more than two years to close this deal.
01:17Yes.
01:17It closed last year.
01:18You've had some time now to start that integration process.
01:21How is it going?
01:22Fantastic, actually.
01:24We released our first wave of products in March with a number of customers.
01:30NVIDIA is one of our key partners in the early adoption of these products, and we're expanding the opportunity to
01:37not only semiconductor customers, to a number of companies that are building those physical systems for the physical AI world.
01:45You've used the phrase, silicon the system.
01:49Yes.
01:49The idea of, is that kind of like farm to table that we hear in restaurants?
01:52It's kind of like the whole ecosystem.
01:54Is that a slogan, or is there a real substance behind that?
01:57It's a real substance when you hear your reference to NVIDIA.
02:01Initially, NVIDIA was talking about chips.
02:04Then it went from chips to rack to the whole data center, now to physical AI.
02:09That's exactly a silicon to a system.
02:12You need to design and optimize the two in the same context in order to reduce the cost, in order
02:19to provide a product that is efficiently built for specific applications.
02:25Some investors have been concerned about when you start to see more meaningful revenue growth tied to this deal.
02:31I know there was some talk about a deceleration in revenue, at least in the second half of this year.
02:36Why?
02:37No, actually, we just reported Q2.
02:39We beat on all metrics.
02:42We raised the year.
02:43We see an expanded opportunity in the future.
02:45But as you know, when you go through a massive integration of two companies, there's a runway before you deliver
02:52the integrated products.
02:54And we have an expectation.
02:55We set with our investors that 2027 will be the year where we start seeing the new products and creating
03:02the new revenue opportunity.
03:04That's right around the corner.
03:05It's right around the corner.
03:06Another part of your business that investors have focused on is the design IP business.
03:12And I know that's a smaller part of your business.
03:14Do you anticipate that that could grow to be a bigger share of your revenue?
03:19Currently, it's about 20% of our business.
03:22I cannot be more excited about this business.
03:24As you imagine the hyperscalers today, they're contemplating three options, buying merchant silicon, doing their ASIC silicon, and building their
03:36own.
03:37If you look at every hyperscaler, they're trying to invest and accelerate their own silicon in data center because of
03:44cost and it's optimized for very specific applications.
03:48That lane will not happen without our IP business.
03:52And I'm not exaggerating that it will not happen without our IP business.
03:56So it's so essential.
03:58The more acceleration you see with system companies building their own silicon, that provides a significant tailwind to our IP
04:06business.
04:07Part of that IP business was for a time dependent on China.
04:11Obviously, there's been a lot of issues there with regards to regulations and the ability or inability to actually sell
04:17there.
04:17Does that improve?
04:18We're assuming China will remain the way it is currently, meaning our China business, specific to IP, was one of
04:27our fastest growing regions that hit a significant headwinds due to all the restrictions.
04:32I don't believe these restrictions are going to change.
04:36If they change, fantastic.
04:37If not, we have plenty of opportunities outside of China to capture the growth.
04:43I am curious about the idea of some of these hyperscalers trying to build their own chips.
04:47Obviously, there's been a lot of discussion about Google's TPUs, and we've seen that several other companies, including Microsoft, Meta,
04:54and others, try to move down this pathway.
04:56Does that journey that these companies are taking, is that going to include Synopsys?
05:03Absolutely.
05:04Because there is no chip today that you can design without Synopsys, period.
05:10So regardless if you are the NVIDIAs of the world doing Merchant or the Broadcom of the world, you're doing
05:15ASIC.
05:15And if you're doing your own chip, you use the Synopsys EDA product, as well as IP, to connect the
05:22chips together and to connect the chip to the outside world.
05:25Does this, though, to a certain extent, threaten your relationship with NVIDIA?
05:30Not at all.
05:31Not at all.
05:32Because actually, we help NVIDIA connect their chip to the ecosystem.
05:37About a year ago, NVIDIA announced one of their key connectors is called NVLink.
05:44And they want to connect the NVIDIA chip to the outside world.
05:48We are connecting the NVLink to the TPU or to other chips.
05:52So, no, we're absolutely part of that ecosystem.
05:55So your partnership with NVIDIA so far is going well?
05:57It's actually going very well.
05:59It has multiple levels.
06:00One of it is at the chip level.
06:02The other one is physical AI-level partnership.
06:05With regards to what investors want, particularly with regards to additional growth,
06:10you recently came to a settlement with Elliott.
06:14You have a new board member, Jesse Cohn, taking over.
06:17Have you had a chance at all to meet with Elliott since that announcement was made last month?
06:21Absolutely.
06:21I was directly involved with Jesse from day one.
06:24And we aligned very quickly on the priority, which is Synopsys has great assets.
06:32There is opportunity to monetize further these assets and improve our profitability,
06:37which we've been talking about publicly the moment we closed the ANSYS acquisition.
06:42So there was an immediate alignment on the priorities.
06:45Any thought, any discussion about additional acquisitions?
06:51We don't talk about these publicly, but we're always looking at what are the solutions needed for our customers
06:58and how can we continue on expanding to deliver a holistic solution to the customer.
07:04Final question here, Sine.
07:05I am curious about just longer term, where you see this going.
07:09I mean, we started off talking about digital intelligence.
07:12Now there's a big talk about physical AI, biological AI is at some point going to become part of the
07:19conversation as well.
07:20If we're sitting here five years from now, what is that conversation like?
07:24You know, there are so many either engineering problems or challenges in general
07:31that scientists and engineers have not been able to solve until there was a revolution at the compute level.
07:39Today, the reason we're able to talk about AI, because of the accelerated compute,
07:43you're able to do things that was not possible before.
07:47In health and life sciences, there are so many discoveries that have not been possible
07:52because of the time it takes.
07:55It may take hundreds of years to simulate, which is not practical.
07:59Now with AI, with quantum on the horizon, there are many, many opportunities to accelerate innovation.
08:05And we will try to reach out the way of working on independent government.
08:05And I will tell you that you can only have a range of global development.
08:06What to do, many people have been able to get it.
Comments